The Bethesda System and Guidelines for theManagement of Cervical Intraepithelial

Neoplasia

The Bethesda System and Guidelines for theManagement of Cervical Intraepithelial

Neoplasia

Attila L Major, MD, PhD

Presentation PlanPresentation Plan

• History of Smear Nomenclature • The Bethesda System• Adequacy of the Specimen• ASCUS, ASC• LSIL• AGUS, AGC, AIS• Algorithm

George N. PapanicolaouNew Cancer Diagnosis (1928)

George N. PapanicolaouNew Cancer Diagnosis (1928)

H. Traut, J.E. AyreH. Traut, J.E. Ayre

SMEAR NOMENCLATURESMEAR NOMENCLATURESMEAR NOMENCLATURE

11

22

33

44

55

NegativeNegative for for malignantmalignant cellscells

InflammatoryInflammatory atypiaatypiaSquamosSquamos atypiaatypiaKoilocytoticKoilocytotic atypiaatypia

Mild Mild dysplasiadysplasiaModerateModerate dysplasiadysplasiaSevereSevere dysplasiadysplasia

CarcinomaCarcinoma in situin situ

Invasive Invasive CarcinomaCarcinoma

NegativeNegative

CIN 1CIN 1CIN 2 CIN 2 CIN 3 CIN 3

CIN 3CIN 3

InvInv CaCa

DDescriptiveescriptive (WHO)(WHO)(1968)(1968)

PapPap(1954)(1954)

CINCIN(1978)(1978)

BethesdaBethesda SystemSystem(1988)(1988)

WithinWithin normal normal limitslimits

ReactiveReactive andand reparativereparative changeschangesASCUSASCUSLSIL; LSIL; includesincludes condylomacondyloma

LSIL; LSIL; includesincludes condylomacondylomaHSILHSILHSILHSIL

HSILHSIL

Invasive Invasive CarcinomaCarcinoma

CinCin IIIIII 20x (20x (exocolexocol), h), h--ALA 0,5%, 75 minALA 0,5%, 75 min

E

LP

E- epithelium, LP-lamina propria

DDéépartements de Pathologie Clinique et de Gynpartements de Pathologie Clinique et de Gynéécologie et Obstcologie et Obstéétrique HUG & EPFL, mai 2002trique HUG & EPFL, mai 2002

Cervical squamous carcinoma precursorsCervical squamous carcinoma precursors

Squamous Intraepithelial Lesion (SIL)Low Grade High Grade

Condyloma Cervical intraepithelial Neoplasia(CIN)Grade 1 Grade 2 Grade 3

Normal Mild Dysplasia Moderate Severe In situDysplasia Dysplasia Carcinoma

Microinvasive Carcinoma

Thickness of HSIL of the cervix

100

80

60

40

20

00.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2

Tissue depth (mm)

Fluo

resc

ence

(% o

f sta

ndar

d) CIN 3

Normal tissue

Cervical stromaEpithelium

Pahernik et al Int J Cancer 78 : 310-314, 1998

Fluorescence image of the cervix after h-ALA applicationFluorescence image of the cervix after h-ALA application

White light Fluorescence

Fluorescence image and white light image of the cervix uteri after the application of 3% acetic acid. Application of 10mg h-ALA in 10ml 0.9% NaCl solution on the cervix during 3 hrs.

The Bethesda System (1988)The Bethesda System (1988)

• The term squamous intraepithelial lesion is characterized by a high spontaneous regression rate and the lack of predictable progression of SIL to invasive cancer

• The Bethesda System replaced 3 levels of CIN with 2 levels, LSIL and HSIL, which could be used to describe any squamous abnormality of the lower genital tract

Summary of the Natural History of Cervical Intraepithelial Neoplasia

Summary of the Natural History of Cervical Intraepithelial Neoplasia

PATIENTS (N) REGRESSION(%)

PERSISTENCE(%)

PROGRESSIONTO CIS

(%)

PROGRESSIONTO INVASION

(%)

CIN I 4504 57 32 11 1

CIN II 2247 43 35 22 5

CIN III 767 32 <56 - >12

CIN, cervical intraepithelial neoplasi; CIS, carcinoma in situ

Östör AG : Natural history of cervical intraepithelial neoplasia : A critical review. Int J Gynecol Pathol 12:186, 1993

Pooled estimates with 95% confidence intervals. ASCUS = atypical squamous cells of undetermined significance;

SIL = squamousintraepithelial lesion.

Pooled estimates with 95% confidence intervals. ASCUS = atypical squamous cells of undetermined significance;

SIL = squamousintraepithelial lesion.

Melnikowintraepithelial1998.

J et al: Natural history of cervical squamouslesions: A meta-analysis. Obstet Gynecol

The Bethesda System 2The Bethesda System 2

• Interobserver and intraobserver variability: Lack of reproducibility for the identification of 3 or 4 categories among different laboratories and even by the same cytologist

• Reducing the discordance between interpretation of cytologic and histologic specimens by placing moderate dysplasia (CIN 2), severe dysplasia(CIN 3) and carcinoma in situ into one category (HSIL)

Adequacy of the Specimen: Unsatisfactory Smear

Adequacy of the Specimen: Unsatisfactory Smear

• TBS 2001 separates unsatisfactory for technical reasons from those processed but unsatisfactory

• Typical rates of unsatisfactory smears are reported as 0.5 % (mean 0.95 %)

• The most common reason of for an unsatisfactory smear is scant cellularity, followed by obscuring inflammation and obscuring blood

• A significant number shows SIL/Ca in follow-up

Adequacy of the specimen: endocervical cellsAdequacy of the specimen: endocervical cells

• In TBS 2001 the absence of endocervical cells does not affect specimen adequacy

• Women with smears lacking endocervical cells are not more likely to have squamous lesions on follow up than women with endocervical cells

• No association between false-negative interpretations of smears and lack of endocervicalcells

• No relevance if the specimen shows HSIL

Algorithm for follow-up of smears unsatisfactory for evaluation

Algorithm for follow-up of smears unsatisfactory for evaluation

American Society of Colposcopy and Cervical Pathology

ASCUSASCUS• Atypical Squamous Cells of Undetermined

Significance• Combination of technical and interpretive

problems• Problems of sampling• Problems of cellular preservation• Problems of microscopic observation• Poor interobserver reproducibility

ASCUS/LSILASCUS/LSIL

• The management of ASCUS/LSIL is of concern because a small but important minority have HSIL or even invasive carcinoma

ASCUS SurveyASCUS Survey

• Incidence of ASCUS: 3 % - 5 %• ASCUS associated with HSIL: 15 %• Almost 50 % of all HSIL is preceded by ASCUS• HPV DNA testing is the most sensitive test for

HSIL (96 % sensitivity)• Negative predictive value of HPV DNA: 99%

Revision Bethesda 2001: ASCUS Revision Bethesda 2001: ASCUS

• Bethesda 1991:– ASCUS-FR: favoring a reactive process– ASCUS-FN: favoring a dsyplastic/neoplastic process– ASCUS-NOS: not other specified

• Bethesda 2001:– ASC-US: undetermined significance– ASC-H: suggestive of HSIL

ASC Survey ASC Survey

• ASC-H (Atypical Squamous Cells favoring HSIL)• ASC-H: 24% - 94% CIN 2, 3 histopathology• ASC-US: 5% - 10% CIN 2, 3 histopathology

HPV DNA Testing for the Management of ASCUS

Author (No of Patients) Repeat CytologySensitivity (95% CI)

HPV TestingSensitivity (95% CI)

Ferris1998 (144) 0.70 (0.42-0.98) 0.89 (0.69-1.00)

Manos 1999 (995) 0.76 (0.65-0.87) 0.89 (0.81-0.97)

Bergeron 2000 (111) 0.67 (0.50-1.00) 0.83 (0.62-1.00)

Lin 2000 (74) NA (not applicable) 1.00

Shlay 2000 NA 0.93 (0.81-1.00)

Salomon 2001 0.85 (0.81-0.89) 0.96 (0.94-0.98)

Human papillomavirus (HPV) load stratified by results of colposcopy and final enrollment histopathology for 2198 womenwith atypical squamous cells of undetermined significance (ASCUS) and 848 women with low-grade squamous intraepitheliallesion (LSIL) enrolled in the ASCUS/LSIL Triage Study (ALTS). Cervical specimens stored in PreservCyt were tested for thepresence of HPV DNA as described previously (6,7). Striped bars represent viral load (pg/mL) for women with cervical smears interpreted as ASCUS by a community pathologist open bars represent the viral load for women with cervical smearsinterpreted as LSIL; upper boundary of the box represents the 75th percentile; lower boundary of the box represents the 25th

percentile; center line in the box represents the median; asterisk represents the mean; whiskers represent 1.5x the distance between the 25% and 75% percentiles; and diamonds represent outliers. CIN = cervical intraepithelial neoplasia.

Sherman M et al: ASCUS/LSIL Triage Study: J Natl Cancer Inst2002

ASC-US ManagementASC-US Management• ASC-US may be the precursor of HSIL before

cancer develops• HPV tests effectively identifies patients at risk for

existent HSIL with an similar efficacy as colposcopy

• Women whose HPV DNA is negative can be similarly managed as negative Pap smears (negative predictive value: 99%)

LSIL SurveyLSIL Survey

• Incidence of LSIL: 1.6 %• LSIL associated with HSIL: 18 % (10 % - 70 %) • 0.3 % inv ca, 68 % CIN 1• Almost 25 % of all HSIL is preceded by LSIL

LSILLSIL

• LSIL: 85% to 90% are high-risk HPV positive• LSIL contains both low-risk and high-risk HPV,

in contrast to HSIL • Predominantly made up of productive viral

infections (condyloma and CIN 1)• TBS combines CIN 1 and HPV changes into

LSIL

LSILLSIL• Factors related to the development: young age, presence

of high-risk HPV, persistence of HPV infection• HPV infection reflects sexual activity with an immature

transformation zone• Only 5.1 % of HPV infection and persistence (6 month)

during the 3 year study developed SIL, 93 % CIN 1 (Ho GYet al, N Engl J Med 1998)

• Progression rate from LSIL to HSIL: 10 % - 20 %

AGUS, AGC and Cervical NeoplasiaAGUS, AGC and Cervical Neoplasia

Atypical Glandular Cells (AGC) is associated with a substantially greater risk for cervical neoplasia than Atypical Squamous Cells (ASC) / LSIL

Ronnett et al, Hum Pathol, 1999

Revision Bethesda 2001: AGUS Revision Bethesda 2001: AGUS • Bethesda 1991:

– AGUS-FR: favoring a reactive process– AGUS-FN: favoring a dsyplastic/neoplastic process– AGUS-NOS: not otherwise specified

• Bethesda 2001:– AGC (endocervical, endometrial) or AGC-NOS (Atypical Glandular Cells

not otherwise specified)– AGC-FN: Atypical Glandular Cells (endocervical or glandular)– Adenocarcinoma in situ (endocervical)

AGC SurveyAGC Survey

• AGC (AGUS-FN) and AGC (NOS) was separated by TBS 2001, because they represent women at different risk for neoplasia

• AGC-FN: 27 % - 96 % CIN 2,3; AIS, invas Ca• AGC-NOS: 9 % - 41 %

AGUS SurveyAGUS Survey

• Incidence of AGUS: 0.1 % - 0.4 %• AGUS associated with HSIL: 9 % - 40 %• AGUS associated with adenoca in situ: 8 %• associated with endometrial ca: 10 %• associated with endometrial hyperplasia: 12 %

AGUS, Significant Lesions and Age

Lesions No of cases (27) Mean Age

AdenoCa (endometr) 12 63

Hyperplasia (endometr) 3 66

CIN 1 4 47

CIN 2, 3 6 45

Metast Ca 1 61

AIS 1 26

Liu et al, J LGTD, 2002

Methods of Histologic Evaluation

Diagnostic technique No. of patients (%)(Total : 88)

Endocervical Curettage (ECC) 31 (35.2 %)

Conisation 5 (5.7 %)

Endometrial Curettage (EMC) 15 (17 %)

ECC / EMC 29 (33 %)

Hysterectomy 6 (6.8 %)

Vaginal Biopsy 2 (2.3 %)

Liu W et al, J LGTD 2002

AGUS after hysterectomyAGUS after hysterectomy

• Because data show that women with glandular cells after hysterectomy rarely developed neoplastic lesions regardless of the history of prior malignancy, TBS 2001 stated that these smears (AGUS) should be called negative

• Origins: prolapse of the fallopian tube, vaginal endometriosis, fistula, vaginal adenosis or metaplasia associated with radiation or chemoth

Management of AGS

FN: endocervical, glandular

Endometrial Curettage

NOS

AGS

EndocervicalCurettage

AGC: NOS and FNAGC: NOS and FN

• AGC (FN) who do not have cervical neoplasiadetected continue to be at risk, management with repeat cervical cytology is unacceptable

• AGC NOS have been found to be at relative low risk for a missed lesion (Veljovich, 1998): repeat cytology (until 4 consecutive negative results)

• Persistent AGC-NOS, AGC-FN or AIS have been found to be high risk: cold-knife conisation

AGC or AIS and Endometrial samplingAGC or AIS and Endometrial sampling

• AGC glandular (atypical endometrial cells)• In all women older than 35• In women younger than 35 who have unexpected

vaginal bleeding

Adenocarcinoma in situ (AIS)Adenocarcinoma in situ (AIS)

• TBS 2001concluded that the criteria for adenocarcinoma in situ have been shown to be predictive and reproducible that a separate category be established

• AIS (cytol) is associated with AIS (histopath) in 48 % - 69 % or invasive cervical adenoca in 38 %

• 50 % have coexisting squamous abnormality

AGC and AIS or AdenocaAGC and AIS or Adenoca

• Many cases of biopsy-confirmed AIS had no observed colposcopic abnormalities, and even combinations with cytology can miss small endocervic adenoca and AIS (Cullimore J, 1992)

• Endocervical sampling can detect glandular neoplasia that is missed by colposcopy

ColpoCEC

Pipelle

3 ansHPV(-)

ColpoCEC

Pipelle

3 ansHPV(-)

ASC-US ASC-H, AGC/LSIL HSIL, AGC-FNAIS

18.10.02

Conclusion (Bethesda System 2001)Conclusion (Bethesda System 2001)• Atypical Squamous Cells (ASC) is subcategorised in ASC-US

(undetermined significance) and in ASC-H (suggestive of HSIL)• ASC-US should be managed using by one of following program:

– cytology follow-up– Immediate colposcopy– High-risk HPV DNA testing

• HPV DNA testing is the preferred approach with liquid-based cyto• No role for HPV testing in LSIL• ASC-H, LSIL, HSIL and AGUS should be referred for immediate

colposcopy• ASC-US with oncogenic HPV types are equated to LSIL